Ship shafting undocking accurate positioning system and adaptive adjustment method

Abstract

The application discloses a ship shafting undocking accurate positioning system and a self-adaptive adjusting method, relates to the technical field of ships, and comprises the following steps: recording a reference ideal state in a rigid support state in a dock; collecting monitoring data reflecting shafting centering state changes in real time in the process that the ship undocks and floats on the water surface; generating adjusting instructions for compensating the errors; and driving multiple self-adaptive adjusting mechanisms arranged at shafting support points to perform collaborative action according to the adjusting instructions, so as to dynamically adjust the spatial positions of the support points and make the shafting keep optimal centering in a ship body deformation environment. The application has the advantages that the shafting centering state changes caused by ship body deformation in the undocking process are sensed in real time by a sensor array, and the self-adaptive adjusting mechanisms are driven to perform dynamic and accurate compensation based on a prediction model and a closed-loop control algorithm, so that the defects of traditional methods, such as dependence on experience and insufficient precision, are effectively overcome.

Description

Technical Field

[0001] This invention relates to the field of marine technology, specifically to a precise positioning system and adaptive adjustment method for ship shafting when leaving dry dock. Background Technology

[0002] As a critical transmission component of the power plant, the alignment accuracy of a ship's shafting directly affects the reliability, efficiency, and vibration and noise levels of its operation. Currently, the installation and alignment of ship shafting are primarily carried out in the dry dock, with the hull under rigid timber support. The alignment determined under these conditions is considered ideal. However, when a ship leaves the dry dock and is launched, transitioning from rigid support to buoyancy support, the hull structure undergoes significant elastic deformation. This disrupts the pre-calibrated shaft alignment in the dry dock, generating additional stresses such as misalignment and bending. Consequently, this leads to a series of problems, including accelerated bearing wear, abnormal vibration, and seal failure, seriously threatening navigational safety.

[0003] To address this problem, existing technologies typically employ methods such as secondary adjustments after undocking or reserving offsets, but these methods have significant limitations. Secondary adjustments after undocking must be performed in a floating state, which is characterized by limited working space, low efficiency, and difficulty in guaranteeing accuracy. Meanwhile, reserving offsets heavily relies on the designer's experience, has poor adaptability to changes in hull shape, and struggles to accurately compensate for complex actual deformations. Therefore, there is an urgent need in the field for an intelligent method and system capable of real-time sensing of hull deformation and dynamic, precise compensation of the shafting alignment during the ship's undocking and floating processes, ensuring that the shafting maintains optimal alignment throughout the entire process from dock to underwater. Summary of the Invention

[0004] To solve the above-mentioned technical problems, a precise positioning system for ship shafting undocking and an adaptive adjustment method are provided. This technical solution solves at least one of the technical problems mentioned in the background section.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0006] A method for adaptive adjustment of a ship's shafting system before undocking includes:

[0007] Under rigid support conditions in the dry dock, the ship's shafting is adjusted to the predetermined theoretical alignment state, and the spatial attitude data of the shafting at this time is recorded as the reference ideal state.

[0008] During the process of a ship leaving the dock and floating on the water, a sensor array deployed on the shafting and hull structure collects monitoring data in real time that reflects changes in the alignment of the shafting.

[0009] Based on the comparison between the monitoring data and the benchmark ideal state, the current alignment error of the shafting is calculated, and combined with the hull deformation prediction model, an adjustment command is generated to compensate for the error.

[0010] According to the adjustment command, multiple adaptive adjustment mechanisms set at the shaft support points are driven to perform coordinated actions to dynamically adjust the spatial position of the support points so that the shaft system maintains optimal alignment under the deformation environment of the hull.

[0011] Preferably, the step of adjusting the ship's shafting to a predetermined theoretical alignment state under rigid support within the dock, and recording the spatial attitude data of the shafting at this time as a reference ideal state, specifically includes:

[0012] A global coordinate system for the hull was established using high-precision geometric measuring instruments, and the theoretical center lines of the axis system were calibrated.

[0013] Control the adaptive adjustment mechanism to precisely adjust each key point of the shaft system to the theoretical center line, and verify that its alignment accuracy meets the predetermined standard;

[0014] Record the readings of each sensor at this time, and construct the initial state of the digital twin of the axis system based on this, which is the reference ideal state.

[0015] Preferably, during the process of the ship leaving the dry dock and floating on the water, the real-time acquisition of monitoring data reflecting changes in the alignment state of the shafting, through sensor arrays deployed on the shafting and hull structure, specifically includes:

[0016] Based on the structural finite element analysis results of the ship's shafting system and the dynamic simulation of the undocking process, the monitoring points on the shafting system are divided into at least three risk levels. Among them, the high-risk level points include at least the stern tube aft bearing housing, the main engine output end support, and the last intermediate bearing support, which are most sensitive to hull deformation.

[0017] Different basic monitoring frequencies and filtering parameters are assigned to locations with different risk levels, and the initial monitoring frequency for high-risk locations is higher than that for medium- and low-risk locations.

[0018] Key events during the undocking process are defined as mode switching triggers, and the key events include at least the start of water injection, ship buoyancy, and initial stabilization of the floating state.

[0019] During the undocking process, the rate of change of sensor readings at each monitoring point relative to the baseline ideal state is calculated in real time. When the rate of change of data at any monitoring point exceeds the first-level safety threshold preset for that point, the system automatically increases the monitoring frequency of that point and its associated points to the first-level high-frequency mode. When the rate of change of data exceeds the higher second-level safety threshold, the system further increases the monitoring mode of that point to the ultra-high frequency mode and simultaneously activates the hull deformation prediction model to perform feedforward compensation calculation.

[0020] When the data change rate of all monitoring points remains below a set stable threshold for a predetermined period of time, the system will automatically adjust the global monitoring frequency from high-frequency mode to a baseline monitoring mode that matches the current undocking phase. However, the monitoring frequency of high-risk points will remain higher than that of medium- and low-risk points.

[0021] All monitoring mode switching events, triggering conditions, and corresponding sensor data streams are recorded throughout the process for subsequent system performance evaluation and algorithm optimization.

[0022] Preferably, the steps for constructing the hull deformation prediction model are as follows:

[0023] Based on the ship's design drawings, a parametric finite element model of the ship's hull beam was established using finite element analysis software. Boundary conditions were applied to the model, where the wooden supports in the dock were simplified to full constraints on the bottom of the hull, and the water in the floating state was simplified to distributed buoyancy applied to the surface of the hull.

[0024] Calculate the deformation of the hull under various typical floating conditions, including at least flat floating, stern trim, bow trim and different heel angle conditions. Extract the theoretical displacement change of the shaft support points under each condition to form a database of the correspondence between "floating parameters and support point displacements".

[0025] During the launching process of the first ship or a ship of the same type, actual floating parameters and support point displacement data measured by a sensor array are collected. The actual data are compared with the theoretical displacement change. Machine learning algorithms are used to correct the theoretical model and train a data-driven deformation prediction proxy model. The floating parameters include draft and trim values.

[0026] The trained agent model is encapsulated into a software module that can be called by the central control system. Inputting real-time floating parameters such as draft and trim, the module can output the predicted displacement of each support point of the shaft system.

[0027] During system use, new measured floating state and deformation data are continuously used as incremental learning samples, and the surrogate model is fine-tuned and optimized periodically so that the prediction accuracy continues to improve with the accumulation of data.

[0028] Preferably, the step of calculating the current alignment error of the shafting system based on the comparison between the monitoring data and the benchmark ideal state, and generating adjustment instructions to compensate for the error in conjunction with the hull deformation prediction model, specifically includes:

[0029] Based on the comparison between the monitoring data and the benchmark ideal state, the current alignment error of the shaft system is calculated as a feedback signal;

[0030] The expected displacement predicted by the hull deformation prediction model is used as a feedforward signal.

[0031] The compensation output is generated by combining the feedback signal and the feedforward signal.

[0032] Preferably, the step of driving multiple adaptive adjustment mechanisms set at the shaft support points to perform coordinated actions according to the adjustment command, and dynamically adjusting the spatial position of the support points so that the shaft system maintains optimal alignment under hull deformation environment specifically includes:

[0033] The central controller receives the overall adjustment command from the intelligent decision-making process. This command contains the target position coordinates or the required position adjustment amount for each shaft support point. The controller parses the overall command into independent sub-commands that correspond one-to-one with each adaptive adjustment support.

[0034] The controller synchronously sends the parsed sub-instructions to each corresponding adaptive adjustment support and sets a unified timing window for the coordinated action of all supports to ensure that the adjustment action of each support starts and progresses synchronously in time, so as to avoid introducing additional bending stress in the shaft system due to asynchronous adjustment.

[0035] Each adaptive adjustment support drives its built-in electric or hydraulic actuator to move according to the received sub-instruction. At the same time, the high-precision displacement sensor built into the support monitors the actual displacement in real time and sends the feedback data back to the central controller. The central controller compares the actual displacement with the target value of the instruction. If there is a deviation, it uses an incremental PID algorithm to generate a fine-tuning instruction to drive the actuator to continue to move until the error between the actual displacement and the target value enters the preset tolerance range, thus forming a closed-loop precision control system.

[0036] After all adaptive adjustment supports have completed precise fine-tuning, the system collects the overall alignment data of the shafting system again through the sensor network of the sensing layer. When it is verified that the alignment status has reached the optimal target, the system issues a prompt signal and remotely controls or manually triggers the mechanical locking device integrated on each adaptive adjustment support to fix the moving parts of the support, so as to form a rigid support that can withstand the long-term operating load of the ship.

[0037] Furthermore, this solution proposes a precise positioning system for ship shafting before undocking, comprising:

[0038] The sensing unit includes a sensor array deployed on the shafting and hull structure, used to collect monitoring data reflecting changes in the alignment status of the shafting in real time during the ship's undocking and floating on the water.

[0039] The reference state storage unit is used to store the spatial attitude data of the shaft system when it is in the theoretical alignment state under the rigid support state of the dock, as the reference ideal state;

[0040] The data processing and decision-making unit is configured to: receive monitoring data sent by the sensing unit, compare it with the benchmark ideal state, calculate the current alignment error of the shaft system, and generate adjustment instructions to compensate for the error in combination with the hull deformation prediction model.

[0041] The execution unit includes multiple adaptive adjustment mechanisms installed at each support point of the shaft system. These mechanisms are used to coordinate actions according to the adjustment commands issued by the data processing and decision-making unit, dynamically adjusting the spatial position of the support points so that the shaft system maintains optimal alignment under hull deformation conditions.

[0042] Optionally, the deployment strategy of the sensor array in the sensing unit is determined based on the structural finite element analysis results of the ship shafting and the dynamic simulation of the undocking process. The sensor array includes at least sensors for monitoring high-risk points, and the high-risk points include at least the stern tube aft bearing seat, the main engine output end support and the last intermediate bearing support, which are most sensitive to hull deformation.

[0043] The sensing unit includes a monitoring mode management module, which is configured as follows:

[0044] Different basic monitoring frequencies and filtering parameters are assigned to locations with different risk levels;

[0045] Define key events during the undocking process as mode switching triggers;

[0046] During the undocking process, the monitoring frequency and mode of the corresponding points are dynamically adjusted based on the comparison results of the rate of change of the sensor readings at each monitoring point relative to the benchmark ideal state and the preset safety threshold.

[0047] Optionally, the data processing and decision-making unit includes an integrated hull deformation prediction proxy model module and an adjustment command calculation module;

[0048] The hull deformation prediction proxy model module is obtained by encapsulating a proxy model based on the finite element theory model and using measured data to correct and train it through machine learning. It is configured to receive real-time floating parameters and output the predicted expected displacement of each support point of the shaft system.

[0049] The adjustment command calculation module is configured to: use the calculated current alignment error as a feedback signal, use the expected displacement output by the hull deformation prediction proxy model module as a feedforward signal, and generate a compensation output adjustment command to drive the execution unit by integrating the feedback control function and the feedforward control function.

[0050] Optionally, each adaptive adjustment mechanism in the execution unit includes an electric or hydraulic actuator, a high-precision displacement sensor, and a mechanical locking device;

[0051] The execution unit includes a central controller, configured as follows:

[0052] The overall adjustment instructions from the data processing and decision-making unit are parsed into independent sub-instructions corresponding to each adaptive adjustment mechanism, and each mechanism is controlled to start and advance adjustment actions within a set unified timing window.

[0053] The system receives the actual displacement from the displacement sensors in each mechanism, compares it with the target value, and uses a closed-loop control algorithm to generate fine-tuning commands until the error between the actual displacement and the target value enters the preset tolerance range.

[0054] After all adaptive adjustment mechanisms have completed precise fine-tuning and the shaft alignment has been verified as qualified, the mechanical locking devices of each mechanism are activated to fix the support.

[0055] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0056] This invention uses a sensor array to monitor changes in shaft alignment caused by hull deformation during the undocking process in real time. Based on a predictive model and closed-loop control algorithm, it drives an adaptive adjustment mechanism for dynamic and precise compensation. This effectively overcomes the shortcomings of traditional methods, which rely on experience and lack precision. It achieves automatic optimal alignment of the shaft system throughout the entire process from dock to underwater, significantly improving shaft installation accuracy and efficiency. It fundamentally avoids problems such as bearing wear and vibration noise caused by misalignment, extending equipment life and ensuring the safety and reliability of ship operation. Simultaneously, this intelligent method reduces reliance on manual experience, lowers secondary operation costs, and provides a significant technological breakthrough for shipbuilding processes. Attached Figure Description

[0057] Figure 1 This is a flowchart of the ship shafting out-of-docking adaptive adjustment method proposed in this invention;

[0058] Figure 2 This is a flowchart of the method for recording the reference ideal state of a shaft system proposed in this invention;

[0059] Figure 3 This is a flowchart of the method for real-time acquisition of monitoring data reflecting changes in the alignment state of the shaft system proposed in this invention;

[0060] Figure 4 This is a flowchart illustrating the method for constructing the ship deformation prediction model proposed in this invention. Detailed Implementation

[0061] The following description is intended to disclose the invention and enable those skilled in the art to implement it. The preferred embodiments described below are merely examples, and other obvious variations will occur to those skilled in the art.

[0062] Reference Figure 1 As shown, a method for adaptive adjustment of a ship's shafting before undocking includes:

[0063] Under rigid support within the dry dock, the ship's shafting is adjusted to a predetermined theoretical alignment state, and the spatial attitude data of the shafting at this point is recorded as the reference ideal state. In the stable, rigid support environment of the dry dock, a high-precision spatial attitude reference is established for the entire shafting. This "reference ideal state" serves as the sole reliable reference for all subsequent dynamic adjustments, fundamentally solving the problem of fuzzy alignment targets caused by hull deformation, and providing a precise "zero" starting point for accurate compensation during the undocking process.

[0064] During the ship's undocking and surfacing processes, sensor arrays deployed on the shafting and hull structure collect real-time monitoring data reflecting changes in the shafting alignment. This dense sensor array enables comprehensive, real-time, and data-driven perception of hull deformation and shafting alignment throughout the entire process. It transforms the traditionally "black box" undocking process into a transparent, quantifiable data stream, shifting from static, experience-based judgment to dynamic, data-driven approaches. This provides a real and continuous data foundation for intelligent decision-making, ensuring the timeliness and accuracy of system responses.

[0065] Based on the comparison between monitoring data and the baseline ideal state, the current alignment error of the shafting is calculated. Combined with the hull deformation prediction model, adjustment commands are generated to compensate for this error. This combines feedback control based on real-time data with feedforward control based on model prediction. Feedback control ensures accurate correction of the current error, while the feedforward model can predictively compensate for impending deformation, thus significantly reducing system response lag. This achieves proactive and smooth adjustment, effectively suppressing excessive dynamic deviations and improving the stability and quality of alignment control.

[0066] According to the adjustment command, multiple adaptive adjustment mechanisms installed at the shaft support points are driven to work in concert, dynamically adjusting the spatial position of the support points to maintain optimal alignment of the shaft system under hull deformation. This translates intelligent decision-making into actual mechanical actions, actively counteracting the effects of hull deformation through the coordinated work of multiple adjustment mechanisms. Its beneficial effect lies in achieving dynamic, closed-loop control of the spatial position of the shaft support points, enabling the shaft centerline to actively adjust to "follow" hull deformation, thus maintaining an ideal straight line state under changing buoyancy conditions. This fundamentally avoids the generation of additional stress and ensures the long-term safe and stable operation of the shaft system.

[0067] Reference Figure 2As shown, under rigid support conditions within the dry dock, the ship's shafting is adjusted to a predetermined theoretical alignment state, and the spatial attitude data of the shafting at this point is recorded as the reference ideal state. Specifically, this includes:

[0068] A global coordinate system for the hull was established using high-precision geometric measuring instruments, and the theoretical center lines of the axis system were calibrated.

[0069] Control the adaptive adjustment mechanism to precisely adjust each key point of the shaft system to the theoretical center line, and verify that its alignment accuracy meets the predetermined standard;

[0070] Record the readings of each sensor at this time, and construct the initial state of the digital twin of the axis system based on this. This initial state is the reference ideal state.

[0071] By establishing a precise global coordinate system and theoretical benchmark, and actively adjusting the axis system to an ideal state using an adaptive adjustment mechanism, the initiative and high precision of benchmark establishment are ensured. This avoids the errors of passively accepting the existing installation state as the benchmark in traditional methods, and establishes a unique, precise, and authoritative basis for subsequent alignment compensation throughout the entire process. Simultaneously, constructing a digital twin initial state ensures that the "ideal benchmark state" is no longer an isolated data point, but a virtual model containing spatial relationships and sensor reading mappings, laying an intelligent foundation for subsequent real-time data comparison and state perception.

[0072] The core of this step lies in establishing a high-precision benchmark through "active construction" rather than "passive measurement." High-precision geometric measurement instruments ensure the acquisition of absolute spatial coordinates; while directly controlling the adjustment mechanism for precise alignment ensures that the shaft system is in its theoretically optimal geometric state when the benchmark is established. Recording sensor readings and constructing the initial state of a digital twin means completely "digitizing" the ideal state of the physical entity. This digital twin will become a virtual reference benchmark and decision engine for real-time judgment of state changes and calculation of adjustment amounts during the undocking process.

[0073] Reference Figure 3 As shown, during the ship's undocking and buoyancy processes, sensor arrays deployed on the shafting and hull structure collect real-time monitoring data reflecting changes in the shafting alignment. Specifically, this includes:

[0074] Based on the structural finite element analysis results of the ship's shafting system and the dynamic simulation of the undocking process, the monitoring points on the shafting system are divided into at least three risk levels. Among them, the high-risk level points include at least the stern tube aft bearing housing, the main engine output end support, and the last intermediate bearing support, which are most sensitive to hull deformation.

[0075] Different basic monitoring frequencies and filtering parameters are assigned to locations with different risk levels, and the initial monitoring frequency for high-risk locations is higher than that for medium- and low-risk locations.

[0076] Key events during the undocking process are defined as mode switching triggers. Key events include at least the start of water injection, ship buoyancy, and initial stabilization of the floating state.

[0077] During the undocking process, the rate of change of sensor readings at each monitoring point relative to the baseline ideal state is calculated in real time. When the rate of change of data at any monitoring point exceeds the first-level safety threshold preset for that point, the system automatically increases the monitoring frequency of that point and its associated points to the first-level high-frequency mode. When the rate of change of data exceeds the higher second-level safety threshold, the system further increases the monitoring mode of that point to the ultra-high frequency mode and simultaneously activates the hull deformation prediction model to perform feedforward compensation calculation.

[0078] When the data change rate of all monitoring points remains below a set stable threshold for a predetermined period of time, the system will automatically adjust the global monitoring frequency from high-frequency mode to a baseline monitoring mode that matches the current undocking phase. However, the monitoring frequency of high-risk points will remain higher than that of medium- and low-risk points.

[0079] All monitoring mode switching events, triggering conditions, and corresponding sensor data streams are recorded throughout the process for subsequent system performance evaluation and algorithm optimization.

[0080] By introducing a differentiated monitoring strategy based on risk level and a dynamic frequency adjustment mechanism based on data change rate, the optimal allocation of monitoring resources and a significant improvement in system response efficiency were achieved. The beneficial effects of this scheme are mainly reflected in three aspects: First, by focusing on monitoring key high-risk points, precise control of the most sensitive and fault-prone areas is ensured under limited resource conditions, effectively preventing major alignment deviations. Second, using key events and data change rate as intelligent triggers, the system can automatically switch to high-frequency or even ultra-high-frequency monitoring modes during critical stages of severe hull deformation and automatically revert to the baseline mode after the state stabilizes. This not only enables agile capture of rapid deformation, gaining valuable time for feedforward compensation, but also avoids the resource waste caused by continuous high-frequency system operation. Finally, the data recording throughout the process provides a valuable data foundation for system performance evaluation and algorithm iteration optimization, enabling the system to continuously improve its self-improvement capabilities and enhance its intelligence level.

[0081] Reference Figure 4 As shown, the steps for constructing the hull deformation prediction model are as follows:

[0082] Based on the ship's design drawings, a parametric finite element model of the ship's hull beam was established using finite element analysis software. Boundary conditions were applied to the model, where the wooden supports in the dock were simplified to full constraints on the bottom of the hull, and the water in the floating state was simplified to distributed buoyancy applied to the surface of the hull.

[0083] The deformation of the hull under various typical floating conditions is calculated. Typical floating conditions include at least flat floating, stern trim, bow trim and different heel angles. The theoretical displacement changes of the shaft support points under each condition are extracted to form a database of the correspondence between "floating parameters and support point displacements".

[0084] During the launching process of the first ship or similar ships, actual floating parameters and support point displacement data measured by sensor arrays are collected. The actual data are compared with the theoretical displacement changes, and machine learning algorithms are used to correct the theoretical model to train a data-driven deformation prediction proxy model. The floating parameters include draft and trim values.

[0085] The trained agent model is encapsulated into a software module that can be called by the central control system. Inputting real-time floating parameters such as draft and trim, the module can output the predicted displacement of each support point of the shaft system.

[0086] During system use, new measured floating state and deformation data are continuously used as incremental learning samples, and the surrogate model is fine-tuned and optimized regularly so that the prediction accuracy continues to improve with the accumulation of data.

[0087] The hull deformation prediction model constructed in this step employs a hybrid modeling approach that integrates physical mechanisms with data-driven methods. This method does not rely entirely on idealized theoretical finite element models, nor is it purely black-box data fitting. Instead, it uses a validated physical theoretical model as its foundation and framework, and then utilizes measured data collected during actual docking to "calibrate" and "refine" the model through machine learning algorithms. This results in a high-precision "surrogate model" that reflects the fundamental physical laws of hull deformation while correcting for errors in the theoretical model, deviations in actual construction, and the influence of environmental factors. This approach effectively combines the generalization ability of theoretical models with the high-precision advantages of data-driven models, making the prediction results closer to engineering reality. Furthermore, by encapsulating the model as a callable module and designing a continuous learning mechanism, the predictive function can be seamlessly integrated into the control system and possess continuous evolution capabilities, ultimately providing core decision-making basis for achieving advanced and precise feedforward compensation control.

[0088] Based on the comparison between monitoring data and the baseline ideal state, the current alignment error of the shafting is calculated. Combined with the hull deformation prediction model, adjustment commands are generated to compensate for this error, specifically including:

[0089] Based on the comparison between the monitoring data and the benchmark ideal state, the current alignment error of the shaft system is calculated as a feedback signal;

[0090] The expected displacement predicted by the hull deformation prediction model is used as a feedforward signal.

[0091] The combined feedback signal and feedforward signal generate a compensated output.

[0092] ;

[0093] in, To compensate for the output, For centering error, For the expected displacement, For feedback control function, It is a feedforward control function;

[0094] The feedback control function is specifically implemented as a variable-gain PID algorithm with dead-time and integral separation, as follows:

[0095] ;

[0096] , , This is a time-varying gain coefficient, whose value is non-linearly adjusted according to the magnitude of |e(t)|: when |e(t)| is large, a more aggressive gain is used to achieve rapid adjustment; when |e(t)| is small, a conservative gain is used to ensure stability. For the proportional term error, when |e(t)| < δ, ,otherwise, This is used to avoid frequent adjustments by the system near the equilibrium point; δ is the dead zone threshold. To address the integral term error, an integral separation strategy is employed. When |e(t)|>E, To prevent saturation, cancel the integral action when |e(t)|≤E. That is, the accumulation of errors, where E is the separation threshold. For differential term error, , For time intervals;

[0097] The feedforward control function is a lead compensator based on the hull deformation prediction model, specifically:

[0098] ;

[0099] The feedforward gain matrix, determined through system identification experiments, is used to match the dynamic response characteristics of the actuator. This is the advance compensation time constant, used to compensate for system response lag. Its value is an estimate of the actuator response delay and filter phase shift. The derivative of the expected displacement represents the deformation trend and is used to provide advance compensation.

[0100] A composite intelligent control algorithm combining improved feedback control and model predictive feedforward control was designed and implemented, bringing several significant benefits. First, the feedback-feedforward composite control architecture enables the system to not only accurately correct the current alignment error but also proactively compensate for impending hull deformation based on the predictive model. This dual "present + future" effect greatly improves the system's response speed and control accuracy, effectively reducing dynamic deviations during transients and achieving smooth adjustment with small overshoot and fast convergence. Second, the variable-gain PID algorithm with dead zone and integral separation used in the feedback control cleverly balances response speed and control stability: rapid adjustment under large deviations and avoidance of oscillations under small deviations or stability; simultaneously, integral separation effectively prevents integral saturation, enhancing the system's stability and reliability. Finally, the derivative term based on deformation trends and advance compensation introduced in the feedforward control can actively offset the system's inherent response lag, allowing for better synchronization between adjustment actions and the hull deformation process, thereby significantly improving the shafting's adaptive capability and robustness in maintaining optimal alignment under dynamic deformation environments.

[0101] According to the adjustment command, multiple adaptive adjustment mechanisms installed at the shaft support points are driven to work in concert to dynamically adjust the spatial position of the support points, so that the shaft system maintains optimal alignment under hull deformation conditions. Specifically, this includes:

[0102] The central controller receives the overall adjustment command from the intelligent decision-making process. This command contains the target position coordinates or the required position adjustment amount for each shaft support point. The controller parses the overall command into independent sub-commands that correspond one-to-one with each adaptive adjustment support.

[0103] The controller synchronously sends the parsed sub-instructions to each corresponding adaptive adjustment support and sets a unified timing window for the coordinated action of all supports to ensure that the adjustment action of each support starts and progresses synchronously in time, so as to avoid introducing additional bending stress in the shaft system due to asynchronous adjustment.

[0104] Each adaptive adjustment support drives its built-in electric or hydraulic actuator to move according to the received sub-instruction. At the same time, the high-precision displacement sensor built into the support monitors the actual displacement in real time and sends the feedback data back to the central controller. The central controller compares the actual displacement with the target value of the instruction. If there is a deviation, it uses an incremental PID algorithm to generate a fine-tuning instruction to drive the actuator to continue to move until the error between the actual displacement and the target value enters the preset tolerance range, thus forming a closed-loop precision control system.

[0105] After all adaptive adjustment supports have completed precise fine-tuning, the system collects the overall alignment data of the shafting system again through the sensor network of the sensing layer. When it is verified that the alignment status has reached the optimal target, the system issues a prompt signal and remotely controls or manually triggers the mechanical locking device integrated on each adaptive adjustment support to fix the moving parts of the support, so as to form a rigid support that can withstand the long-term operating load of the ship.

[0106] This system achieves a triple guarantee of "coordinated synchronization," "precise closed-loop," and "final rigid locking." First, through command parsing and synchronous timing control by the central controller, the actions of multiple adjustment mechanisms are ensured to be uniform in time. This avoids secondary bending damage to the shafting system that might be caused by step-by-step, asynchronous adjustments, and is key to achieving "harmless" dynamic adjustment. Second, the "local" position closed loop formed within each support (based on high-precision displacement sensors) and the "global" alignment state closed loop formed at the system level (based on sensor network verification) constitute a dual-closed-loop control system, thus ensuring extremely high execution accuracy at both the macro-adjustment and micro-positioning levels. Finally, the introduction of a mechanical locking device is crucial. It allows the adaptive adjustment mechanism to transition from an adjustable "actuator" state to a reliable "structural support" state after completing precise positioning, ensuring that the shafting support possesses the same mechanical reliability as traditional rigid supports during long-term ship operation, achieving a balance between temporary precise adjustment and long-term safe load-bearing.

[0107] Furthermore, the present invention also proposes a ship shafting undocking precision positioning system, comprising:

[0108] The sensing unit includes a sensor array deployed on the shafting and hull structure, used to collect monitoring data reflecting changes in the alignment status of the shafting in real time during the ship's undocking and floating on the water.

[0109] The reference state storage unit is used to store the spatial attitude data of the shaft system when it is in the theoretical alignment state under the rigid support state of the dock, as the reference ideal state;

[0110] The data processing and decision-making unit is configured to: receive monitoring data sent by the sensing unit, compare it with the benchmark ideal state, calculate the current alignment error of the shaft system, and generate adjustment instructions to compensate for the error in combination with the hull deformation prediction model.

[0111] The execution unit includes multiple adaptive adjustment mechanisms set at each support point of the shaft system. These mechanisms are used to coordinate actions according to the adjustment commands issued by the data processing and decision-making unit, dynamically adjust the spatial position of the support points, and ensure that the shaft system maintains optimal alignment under hull deformation conditions.

[0112] The deployment strategy of the sensor array in the sensing unit is determined based on the structural finite element analysis results of the ship shafting and the dynamic simulation of the undocking process. The sensor array includes at least sensors for monitoring high-risk points. High-risk points include at least the stern tube aft bearing housing, main engine output end support and last intermediate bearing support, which are most sensitive to hull deformation.

[0113] The sensing unit includes a monitoring mode management module, which is configured as follows:

[0114] Different basic monitoring frequencies and filtering parameters are assigned to locations with different risk levels;

[0115] Define key events during the undocking process as mode switching triggers;

[0116] During the undocking process, the monitoring frequency and mode of the corresponding points are dynamically adjusted based on the comparison results of the rate of change of the sensor readings at each monitoring point relative to the benchmark ideal state and the preset safety threshold.

[0117] The data processing and decision-making unit includes an integrated hull deformation prediction proxy model module and a regulation command calculation module;

[0118] The hull deformation prediction proxy model module is obtained by encapsulating a proxy model based on the finite element theory model and using measured data to correct and train it through machine learning. It is configured to receive real-time floating parameters and output the predicted expected displacement of each support point of the shaft system.

[0119] The adjustment command calculation module is configured to: use the calculated current alignment error as a feedback signal, use the expected displacement output by the hull deformation prediction proxy model module as a feedforward signal, and generate a compensation output adjustment command for driving the execution unit by integrating the feedback control function and the feedforward control function.

[0120] Each adaptive adjustment mechanism in the actuator unit includes an electric or hydraulic actuator, a high-precision displacement sensor, and a mechanical locking device;

[0121] The execution unit includes a central controller, which is configured as follows:

[0122] The overall adjustment instructions from the data processing and decision-making unit are parsed into independent sub-instructions corresponding to each adaptive adjustment mechanism, and each mechanism is controlled to start and advance adjustment actions within a set unified timing window.

[0123] The system receives the actual displacement from the displacement sensors in each mechanism, compares it with the target value, and uses a closed-loop control algorithm to generate fine-tuning commands until the error between the actual displacement and the target value enters the preset tolerance range.

[0124] After all adaptive adjustment mechanisms have completed precise fine-tuning and the shaft alignment has been verified as qualified, the mechanical locking devices of each mechanism are activated to fix the support.

[0125] In summary, the advantages of this invention are as follows: By using a sensor array to perceive changes in the shaft alignment state caused by hull deformation during the undocking process in real time, and based on a predictive model and closed-loop control algorithm, the adaptive adjustment mechanism is driven to perform dynamic and precise compensation. This effectively overcomes the shortcomings of traditional methods, which rely on experience and lack precision. It achieves automatic optimal alignment of the shaft system throughout the entire process from dock to underwater, significantly improving shaft installation accuracy and efficiency. It fundamentally avoids problems such as bearing wear and vibration noise caused by poor alignment, extends equipment life, and ensures the safety and reliability of ship operation. Simultaneously, this intelligent method reduces reliance on manual experience, lowers secondary operation costs, and provides a significant technological breakthrough for shipbuilding processes.

[0126] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the claimed invention. The scope of protection claimed by the appended claims and their equivalents is defined.

Claims

1. A method for adaptive adjustment of a ship's shafting system before undocking, characterized in that, include: Under rigid support conditions in the dry dock, the ship's shafting is adjusted to the predetermined theoretical alignment state, and the spatial attitude data of the shafting at this time is recorded as the reference ideal state. During the process of a ship leaving the dock and floating on the water, a sensor array deployed on the shafting and hull structure collects monitoring data in real time that reflects changes in the alignment of the shafting. Based on the comparison between the monitoring data and the benchmark ideal state, the current alignment error of the shafting is calculated, and combined with the hull deformation prediction model, an adjustment command is generated to compensate for the error. According to the adjustment command, multiple adaptive adjustment mechanisms set at the shaft support points are driven to perform coordinated actions to dynamically adjust the spatial position of the support points so that the shaft system maintains optimal alignment under the deformation environment of the hull.

2. The method for adaptive adjustment of a ship shafting system before docking according to claim 1, characterized in that, The process of adjusting the ship's shafting to a predetermined theoretical alignment state under rigid support within the dock, and recording the spatial attitude data of the shafting at this point as a reference ideal state, specifically includes: A global coordinate system for the hull was established using high-precision geometric measuring instruments, and the theoretical center lines of the axis system were calibrated. Control the adaptive adjustment mechanism to precisely adjust each key point of the shaft system to the theoretical center line, and verify that its alignment accuracy meets the predetermined standard; Record the readings of each sensor at this time, and construct the initial state of the digital twin of the axis system based on this, which is the reference ideal state.

3. The method for adaptive adjustment of a ship shafting system before docking according to claim 2, characterized in that, The process of launching the ship from the dry dock and floating on the water, specifically the real-time acquisition of monitoring data reflecting changes in the alignment state of the shafting through sensor arrays deployed on the shafting and hull structure, includes: Based on the structural finite element analysis results of the ship's shafting system and the dynamic simulation of the undocking process, the monitoring points on the shafting system are divided into at least three risk levels. Among them, the high-risk level points include at least the stern tube aft bearing housing, the main engine output end support, and the last intermediate bearing support, which are most sensitive to hull deformation. Different basic monitoring frequencies and filtering parameters are assigned to locations with different risk levels, and the initial monitoring frequency for high-risk locations is higher than that for medium- and low-risk locations. Key events during the undocking process are defined as mode switching triggers, and the key events include at least the start of water injection, ship buoyancy, and initial stabilization of the floating state. During the undocking process, the rate of change of sensor readings at each monitoring point relative to the baseline ideal state is calculated in real time. When the rate of change of data at any monitoring point exceeds the first-level safety threshold preset for that point, the system automatically increases the monitoring frequency of that point and its associated points to the first-level high-frequency mode. When the rate of change of data exceeds the higher second-level safety threshold, the system further increases the monitoring mode of that point to the ultra-high frequency mode and simultaneously activates the hull deformation prediction model to perform feedforward compensation calculation. When the data change rate of all monitoring points remains below a set stable threshold for a predetermined period of time, the system will automatically adjust the global monitoring frequency from high-frequency mode to a baseline monitoring mode that matches the current undocking phase. However, the monitoring frequency of high-risk points will remain higher than that of medium- and low-risk points. All monitoring mode switching events, triggering conditions, and corresponding sensor data streams are recorded throughout the process for subsequent system performance evaluation and algorithm optimization.

4. The method for adaptive adjustment of a ship shafting system before docking according to claim 3, characterized in that, The steps for constructing the hull deformation prediction model are as follows: Based on the ship's design drawings, a parametric finite element model of the ship's hull beam was established using finite element analysis software. Boundary conditions were applied to the model, where the wooden supports in the dock were simplified to full constraints on the bottom of the hull, and the water in the floating state was simplified to distributed buoyancy applied to the surface of the hull. Calculate the deformation of the hull under various typical floating conditions, including at least flat floating, stern trim, bow trim and different heel angles. Extract the theoretical displacement change of the shaft support points under each condition to form a database of the correspondence between "floating parameters and support point displacements". During the launching process of the first ship or a ship of the same type, actual floating parameters and support point displacement data measured by a sensor array are collected. The actual data are compared with the theoretical displacement change. Machine learning algorithms are used to correct the theoretical model and train a data-driven deformation prediction proxy model. The floating parameters include draft and trim values. The trained agent model is encapsulated into a software module that can be called by the central control system. Inputting real-time floating parameters such as draft and trim, the module can output the predicted displacement of each support point of the shaft system. During system use, new measured floating state and deformation data are continuously used as incremental learning samples, and the surrogate model is fine-tuned and optimized periodically so that the prediction accuracy continues to improve with the accumulation of data.

5. The method for adaptive adjustment of a ship shafting system before docking according to claim 4, characterized in that, The process of calculating the current alignment error of the shafting system based on the comparison between the monitoring data and the benchmark ideal state, and generating adjustment instructions to compensate for the error in conjunction with the hull deformation prediction model, specifically includes: Based on the comparison between the monitoring data and the benchmark ideal state, the current alignment error of the shaft system is calculated as a feedback signal; The expected displacement predicted by the hull deformation prediction model is used as a feedforward signal. The compensation output is generated by combining the feedback signal and the feedforward signal.

6. The method for adaptive adjustment of a ship shafting system before docking according to claim 5, characterized in that, The step of driving multiple adaptive adjustment mechanisms installed at the shaft support points to perform coordinated actions according to the adjustment command, and dynamically adjusting the spatial position of the support points so that the shaft system maintains optimal alignment under hull deformation conditions, specifically includes: The central controller receives the overall adjustment command from the intelligent decision-making process. This command contains the target position coordinates or the required position adjustment amount for each shaft support point. The controller parses the overall command into independent sub-commands that correspond one-to-one with each adaptive adjustment support. The controller synchronously sends the parsed sub-instructions to each corresponding adaptive adjustment support and sets a unified timing window for the coordinated action of all supports to ensure that the adjustment action of each support starts and progresses synchronously in time, so as to avoid introducing additional bending stress in the shaft system due to asynchronous adjustment. Each adaptive adjustment support drives its built-in electric or hydraulic actuator to move according to the received sub-instruction. At the same time, the high-precision displacement sensor built into the support monitors the actual displacement in real time and sends the feedback data back to the central controller. The central controller compares the actual displacement with the target value of the instruction. If there is a deviation, it uses an incremental PID algorithm to generate a fine-tuning instruction to drive the actuator to continue to move until the error between the actual displacement and the target value enters the preset tolerance range, thus forming a closed-loop precision control system. After all adaptive adjustment supports have completed precise fine-tuning, the system collects the overall alignment data of the shafting system again through the sensor network of the sensing layer. When it is verified that the alignment status has reached the optimal target, the system issues a prompt signal and remotely controls or manually triggers the mechanical locking device integrated on each adaptive adjustment support to fix the moving parts of the support, so as to form a rigid support that can withstand the long-term operating load of the ship.

7. A precise positioning system for ship shafting undocking, characterized in that, include: The sensing unit includes a sensor array deployed on the shafting and hull structure, used to collect monitoring data reflecting changes in the alignment status of the shafting in real time during the ship's undocking and floating on the water. The reference state storage unit is used to store the spatial attitude data of the shaft system when it is in the theoretical alignment state under the rigid support state of the dock, as the reference ideal state; The data processing and decision-making unit is configured to: receive monitoring data sent by the sensing unit, compare it with the benchmark ideal state, calculate the current alignment error of the shaft system, and generate adjustment instructions to compensate for the error in combination with the hull deformation prediction model. The execution unit includes multiple adaptive adjustment mechanisms installed at each support point of the shaft system. These mechanisms are used to coordinate actions according to the adjustment commands issued by the data processing and decision-making unit, dynamically adjusting the spatial position of the support points so that the shaft system maintains optimal alignment under hull deformation conditions.

8. A ship shafting undocking precision positioning system according to claim 7, characterized in that, The deployment strategy of the sensor array in the sensing unit is determined based on the structural finite element analysis results of the ship shafting and the dynamic simulation of the undocking process. The sensor array includes at least sensors for monitoring high-risk points, which include at least the stern tube aft bearing housing, the main engine output end support and the last intermediate bearing support, which are most sensitive to hull deformation. The sensing unit includes a monitoring mode management module, which is configured as follows: Different basic monitoring frequencies and filtering parameters are assigned to locations with different risk levels; Define key events during the undocking process as mode switching triggers; During the undocking process, the monitoring frequency and mode of the corresponding points are dynamically adjusted based on the comparison results of the rate of change of the sensor readings at each monitoring point relative to the benchmark ideal state and the preset safety threshold.

9. A ship shafting undocking precision positioning system according to claim 7, characterized in that, The data processing and decision-making unit includes an integrated hull deformation prediction proxy model module and an adjustment command calculation module. The hull deformation prediction proxy model module is obtained by encapsulating a proxy model based on the finite element theory model and using measured data to correct and train it through machine learning. It is configured to receive real-time floating parameters and output the predicted expected displacement of each support point of the shaft system. The adjustment command calculation module is configured to: use the calculated current alignment error as a feedback signal, use the expected displacement output by the hull deformation prediction proxy model module as a feedforward signal, and generate a compensation output adjustment command to drive the execution unit by integrating the feedback control function and the feedforward control function.

10. A ship shafting undocking precision positioning system according to claim 7, characterized in that, Each adaptive adjustment mechanism in the execution unit includes an electric or hydraulic actuator, a high-precision displacement sensor, and a mechanical locking device. The execution unit includes a central controller, configured as follows: The overall adjustment instructions from the data processing and decision-making unit are parsed into independent sub-instructions corresponding to each adaptive adjustment mechanism, and each mechanism is controlled to start and advance adjustment actions within a set unified timing window. The system receives the actual displacement from the displacement sensors in each mechanism, compares it with the target value, and uses a closed-loop control algorithm to generate fine-tuning commands until the error between the actual displacement and the target value enters the preset tolerance range. After all adaptive adjustment mechanisms have completed precise fine-tuning and the shaft alignment has been verified as qualified, the mechanical locking devices of each mechanism are activated to fix the support.

Images